Download Chapter 4 - DiVA portal

Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Science and Technology 631
_____________________________
_____________________________
Electron and Energy Transfer in
Supramolecular Complexes Designed for
Artificial Photosynthesis
BY
HELENA BERGLUND BAUDIN
ACTA UNIVERSITATIS UPSALIENSIS
UPPSALA 2001
Dissertation for the Degree of Doctor of Philosophy in Physical Chemistry presented at
Uppsala University in 2001.
ABSTRACT
Berglund Baudin, H. 2001. Electron and Energy Transfer in Supramolecular Complexes
Designed for Artificial Photosynthesis. Acta Universitatis Upsaliensis. Comprehensive
Summaries of Uppsala Dissertations from the Faculty of Science and Technology 631. 50 pp.
Uppsala. ISBN 91-554-5033-4
In the society of today the need for alternative energy sources is increasing. The construction
of artificial devices for the conversion of sunlight into electricity or fuel seems very attractive
from an environmental point of view, since these devices are based on processes that does not
necessarily generate any harmful biproducts. In the oxygen evolving photosynthetic process
highly efficient energy and electron transfer reactions are responsible for the conversion of the
sunlight into chemically stored energy and if the same principles can be used in an artificial
device, the only electron supply required, is water.
This thesis describes energy and electron transfer reactions in supramolecular complexes
where the reactions are intended to mimic the basic steps in the photosynthetic process. All
complexes are based on ruthenium(II)-trisbipyridine as photosensitizer, that is covalently
linked to electron donors or electron or energy acceptors. The photochemical reactions were
studied with time resolved transient absorption and emission measurements. In the complexes
that mimic the donor side of Photosystem II, where a manganese cluster together with
tyrosine catalyses the oxidation of water, intramolecular electron transfer was found to occur
from Mn(II) or tyrosine to photo-oxidized Ru(III). Studies of a series of Ru(II)-Mn(II)
complexes gave information of the quenching of the Ru(II) excited state by the coordinated
Mn(II), which is important for the development of multi-nuclear Ru(II)-Mn complexes. In the
supramolecular triad, PTZ-Ru2+-Q, the charge separated state, PTZ+•-Ru2+-Q−•, was rapidly
formed, and further development where a second electron acceptor is linked to quinone is
planned. Ultra fast energy transfer (τ<200 fs), was obtained between ruthenium(II) and
osmium(II) in a small artificial antenna fragment. Fast and efficient energy transfer is
important in larger antennas or photonic wires where a rapid energy transfer is desired over a
large distance.
Key words: Artificial photosynthesis, electron transfer, energy transfer, ruthenium,
manganese.
Helena Berglund Baudin, Department of Physical Chemistry, Uppsala University, Box 532,
S- 751 21 Uppsala, Sweden
©Helena Berglund Baudin 2001
ISSN 1104-232X
ISBN 91-554-5033-4
Printed in Sweden by Akademitryck AB, Edsbruk, 2001
List of papers
This thesis is based on the following papers, which will be referred to in the text
by their Roman numbers.
I
Intramolecular
Electron
Transfer
from
Manganese(II)
Coordinatively Linked to a Photogenerated Ru(III)-Polypyridine
Complex: A Kinetic Analysis
H. Berglund Baudin, L. Sun, R. Davidov, M. Sundahl, S. Styring, B. Åkermark, M.
Almgren, and L. Hammarström
J. Phys. Chem. A, 1998, 102, 2512.
II
Mimicking Electron Transfer Reactions in Photosystem II: Synthesis
and Photochemical Characterization of a Ruthenium(II)
Tris(bipyridyl)Complex with a Covalently Linked Tyrosine
A. Magnuson, H. Berglund, P. Korall, L. Hammarström, B. Åkermark, S. Styring, and
L. Sun
J. Am. Chem. Soc. 1997, 119, 10720.
III
Ruthenium-Manganese Complexes for Artificial Photosynthesis:
Factors Controlling Electron Transfer and Excited State Quenching
Reactions
M. Abrahamsson, H. Berglund Baudin, A. Tran, C. Philouze, K. Berg, M. K.
Raymond- Johansson, L. Sun, B. Åkermark, S. Styring, and L. Hammarström
Manuscript to be submitted to Inorg. Chem.
IV
Efficient Light-induced Charge Separation in a PhenothiazineRu(bpy)32+-Quinone Triad
H. Berglund Baudin, O. Johansson, R. Lomoth, M. Borgström, S. Wallin, L. Sun, B.
Åkermark, and L. Hammarström
In manuscript
V
Ultrafast Energy
Antennas
Transfer
in
Artificial
Ruthenium-Osmium
H. Berglund Baudin, J. Davidsson, S. Serroni, A. Juris, V. Balzani, S. Campagna, and
L. Hammarström
Manuscript to be submitted to J. Phys. Chem.
Reprints were made with permission from the publishers.
Comments on my participation
I carry the main responsibility for all experimental work, data analysis
and manuscripts in paper I and V. In paper II, I did all photophysical and
transient absorption measurements and wrote parts of the manuscript. In
paper III, I performed the emission and electron transfer studies in some
of the complexes and in paper IV, I was responsible for all femtosecond
transient absorption measurements.
Contents
1
Introduction
7
2
Photoinduced Processes of Ru(bpy)32+
10
Properties of Ru(bpy)32+
2.1.1 The ground state
2.1.2 The excited state
Excited state reactions
2.2.1 Photoinduced electron transfer
2.2.2 Photoinduced energy transfer
Time resolved studies of electron and energy transfer
10
10
12
14
15
16
18
Mimicking the Electron Transfer Reactions of Photosystem II
20
3.1
20
20
22
26
29
2.1
2.2
2.3
3
3.2
3.3
4
Mimicking the Light Harvesting Process - Artificial antennas
4.1
4.2
4.3
A
Electron transfer from linked Mn(II) to photo-oxidized Ru(III)
3.1.1 Quenching of the excited of Ru(II) by Mn(II)
3.1.2 Electron transfer from Mn(II) to photo-oxidized Ru(III)
Electron transfer from tyrosine to photo-oxidized Ru(III)
Electron transfer in a covalently linked triad
33
Artificial antennas based on transition metal complexes
33
Ultra fast energy transfer between Ru(II) and Os(II) in a small 36
artificial antenna
Comments on a possible energy transfer mechanism
41
Experimental
43
A.1
A.2
43
45
Femtosecond pump-probe measurements
Nanosecond transient absorption measurements
Acknowledgements
46
Bibliography
47
Chapter 1
Introduction
In the photosynthetic process in green plants, algae and cyanobacteria, sunlight
is absorbed and converted into chemical energy [1]. The reactions responsible
for this process is a series of highly efficient energy and electron transfer steps
that results in the formation of energy rich compounds from the reduction of
carbondioxide. The electron supply for this process is the oxidation of water that
releases oxygen, necessary to sustain life on earth. Since the need for alternative
energy sources is increasing in the society of today, much research is devoted to
the construction of artificial systems that are capable of absorbing sunlight and
convert the energy into electricity or fuel. Systems based on the same principles
as in nature are very attractive since the conversion of sunlight into chemically
stored energy does not necessarily generate any harmful bi-products, and the
only electron source needed is water.
The photosynthetic machinery is located in the thylakoid membrane inside the
chloroplasts [1]. The reaction centers Photosystem II (PSII) and Photosystem I
(PS I), are large protein complexes positioned in the membrane. To increase the
light harvesting efficiency the reaction centers are surrounded by many
chlorophyll molecules, capable of absorbing the sunlight, that are arranged in
large light harvesting complexes. The structural organization of the chlorophyll
pigments within the protein results in an energetically downhill process from the
peripheral antenna complexes to the reaction center. The details of the structure
and energy transfer process are less well known for oxygen evolving species
than in photosynthetic purple bacteria, where the whole process from light
absorption to excitation of the reaction center chlorophylls is typically finished
within approximately 100 ps [2]. The energy transfer rate between individual
pigments within the light harvesting complexes is much faster, and the rate
limiting step is believed to be the energy transfer from the core light-harvesting
complex I to the reaction center.
In PSII where the oxidation of water occurs, the excitation energy harvested by
the antenna complexes eventually leads to excitation of P680, that is the primary
electron donor chlorophyll(s) in the reaction center. The excited P680* then
7
transfers an electron to the acceptor pheophytin and then further to two
quinones, QA and QB. To stabilize the charge separated state the oxidized P680+
is rapidly regenerated by electron transfer from a tyrosyl residue that is
positioned between P680 and the manganese cluster that catalyses the oxidation
of water [1b,3]. Recent publications have suggested that tyrosine is directly
involved in the oxidation of water through a hydrogen atom transfer from a
water molecule to the deprotonated tyrosine [4,5]. Four electron transfer cycles
eventually leads to oxidation of two water molecules coordinated to the
manganese cluster releasing one molecule of oxygen.
e-
e-
P*680
O2 +
e-
Ru2+*
e-
Pheo
QA
4H+
QB
hν
hν
Mn4
e-
e-
eTyrz
2H2O
A
P680
Tyr/Mn
Ru2+
Figure 1.1 Natural (left) versus artificial photosynthesis (right).
Much work is devoted to the understanding of the basic processes of
photosynthesis in natural systems and a lot of effort is also devoted to
construction of complexes that are able to mimic these processes. Manganese
complexes have been constructed as structural mimics of the oxygen evolving
center but these have in general not been coupled to light driven reactions
[3b,6]. There are several mimics of the primary charge separation process
[7,8,9,10,11] but transfer of more than one electron [12] or electron transfer
coupled to drive chemical reactions [13] are unusual. Fast and efficient energy
transfer over large distances, are interesting for the construction of artificial
antennas or photonic wires. Antennas based on porphyrins have been reported
where the excitation energy is efficiently transferred to the unit with lowest
excited state energy [14,15], recently an artificial antenna have also been
coupled to charge separation [16]. Artificial antennas based on transition metal
complexes have been constructed where the synthetic procedure allows a control
of the exact position of each metal unit [17]. Complexes with as much as 22
metal centers have been synthesized where the absorption in the whole UV
8
visible range is large [18]. In these complexes the excitation energy is efficiently
transferred to the unit with lowest excited state energy [19].
In this thesis, work is presented that concern electron and energy transfer
reactions in complexes where the reactions are intended to mimic the basic steps
in natural photosynthesis. All complexes are based on Ru(bpy)32+ as
photosensitizer and depending on the nature of the linked compound different
reactions involving the Ru(bpy)32+excited state will occur. In chapter 2 the basic
properties of Ru(II)-polypyridine complexes in the ground and excited state are
described, and the theories used for electron and energy transfer reactions are
briefly presented. Chapter 3 is devoted to electron transfer reactions in
complexes that are constructed to mimic the electron transfer reactions on the
donor side of PS II. Electron transfer in ruthenium complexes linked with either
a tyrosine or manganese ion are presented (Paper I, II, III). The results obtained
in these studies where an electron was transferred from tyrosine or Mn(II) to
photo-oxidized Ru(III) were the first steps towards mimicking the reactions on
the donor side of PS(II), where P680+ is reduced by electron transfer from
tyrosine. Factors that govern the electron transfer and the energy wasting,
quenching of the excited state of the sensitizer, are discussed. In the last section
of chapter 3 some preliminary results on electron transfer in a covalently linked
triad are presented, intended for development towards complexes that can be
capable of two-electron processes, eventually used for fuel production (Paper
IV). The properties of artificial antennas based on transition metal complexes
are presented in chapter 4 that also contains results on ultra fast energy transfer
in a dinuclear Ru(II)-Os(II) complex (Paper V).
9
Chapter 2
Photoinduced Processes of Ru(bpy)32+
Ruthenium(II)-trisbipyridine has been used in numerous investigations as a
photosenzitizer during the last 30 years due to the very favorable photochemical
properties [20,21].The absorbance in both the visible and UV regions is high. In
the excited state, Ru(bpy)32+ is both a good reductant and oxidant, and the
lifetime is long enough to be used in bimolecular electron or energy transfer
reactions. In addition both the reduced and oxidized forms are relatively stable
towards degrading reactions.
This chapter is intended to give a brief overview of Ru(bpy)32+ as a
photosensitizer. In the first section the properties of the ground and excited
states are presented, followed by a description of electron and energy transfer
processes involving the excited state, and for a more detailed overview of the
properties of Ru(bpy)32+ there are several review articles and books describing
its photophysical and photochemical properties [22,23].
2.1 Properties
2.1.1 The ground state
Ru(bpy)32+ is a d6 transition metal complex with octrahedral geometry. A
simplified molecular orbital diagram of Ru(bpy)32+ is shown in Figure 2.1
together with the ground state absorption spectrum. When transition metals like
ruthenium form complexes with polypyridine ligands the molecular orbitals
created will be localized either predominantly on the metal or on the ligands.
Several absorption bands appear in the ground state absorption spectrum that
correspond to different types of transitions between the molecular orbitals in the
complex. Promotion of an electron from the metal to the ligands is a metal-toligand charge transfer (MLCT) transition, whereas promotion of an electron
from the ligand to the metal is a ligand-to-metal charge transfer (LMCT)
transition. In addition, both metal centered (MC) transitions between orbitals
mainly localized on the metal, and ligand centered (LC) transitions between
orbitals mainly localized on the ligands, occur. Which transition that will be
lowest in energy depends on the oxidation state of the metal and on the nature of
10
the ligands, but for most Ru(II)-complexes with polypyridine type ligands the
transition of lowest energy is the 1MLCT transition [20]. The corresponding
lowest excited states are the 3MLCT states.
The ground state electronic configuration of Ru(bpy)32+ is a singlet state with the
six valence electrons in the t2g (πM) orbitals. When Ru(bpy)32+ is excited at 450
nm, which is the maximum of the band corresponding to the 1MLCT transition,
an electron is transferred to the π* orbital of the ligands, which is the lowest
unoccupied molecular orbital. The ligand centered transitions (LC) of
Ru(bpy)32+, which is transfer of an electron from a π-orbital to π*-orbital, both
localized on the ligand, corresponds to absorption bands in the UV. Metal
centered (MC) transitions give rise to a weak absorption band appearing as a
shoulder at 350 nm. The LMCT transitions for Ru(bpy)32+ are high in energy and
not visible in the absorption spectrum, but for the oxidized form Ru(bpy)33+, the
LMCT state will be the lowest excited state [21]. In Os(bpy)32+ the
corresponding transitions are at lower energy and due to the enhanced spin-orbit
coupling the absorption directly to the triplet states is visible in the spectrum
[21].
LC
!2+
N
1,5
Abs.
N
eg
πL*
N
Ru
N
σM *
N
N
πM
1,0
πL
t2g
MC MLCT
LC
σL
0,5
MLCT
MC
0,0
300
400
500
600
wavelength/nm
Figure 2.1 Absorption spectrum of Ru(bpy)32+(in acetonitrile) together with a schematic
molecular orbital diagram. For explanations see text.
11
2.1.2 The excited state
When Ru(bpy)32+ is excited in the absorption band corresponding to the 1MLCT
transition, an electron is promoted from the metal to the ligand(s). The direction
of the transition dipole moment for the MLCT transition is along the metal to
ligand axis. The excited state is believed to be localized on one bipyridine at a
time [24] and the hopping rate between the bipyridine ligands of the excited
electron have been reported to occur on the picosecond time scale for Ru(bpy)32+
in acetonitrile [25,26]. However a recent investigation has indicated that the
electron is delocalized over all three ligands initially but that the localization of
the electron to one of the bipyridines is ultra fast (τ≈60 fs) [27]. The lowest
excited state of Ru(bpy)32+ is the triplet MLCT state and the emission occurs
from this state which is believed to consist of three levels which are close in
energy [28,29]. Due to the enhanced spin-orbit coupling induced by ruthenium,
the conversion from the 1MLCT to 3MLCT is rapid, τ≈100 fs [30]. The quantum
yield for the formation of the 3MLCT state(s) is near unity [31]. The
deactivation of the 3MLCT state is temperature dependent, and at room
temperature the lifetime of Ru(bpy)32+ is 0.6 µs and the quantum yield of
emission is 0.04 (water). At 77 K the lifetime is increased to 5 µs and the
quantum yield to 0.4 [21].
The lifetime of the 3MLCT state is given by:
1/τ=kr+knr+kd-d
(1)
where kr is the radiative rate constant, knr is the non radiative rate constant and
kd-d is the rate constant for population of the thermally accessible MC state
which decays rapidly to the ground state without radiation. The temperature
dependent process is mainly the decay via the thermally activated MC state,
which reduces the lifetime of the emission at higher temperatures. The nonradiative rate constant, knr is dependent on the energy gap between the ground
state and the 3MLCT state according to the energy gap law [32]. A smaller
energy gap between the lowest excited state and the ground state increases the
non radiative rate constant which makes the excited state lifetime shorter. At
room temperature the lifetime of Ru(bpy)32+ is still long enough for the
excitation energy to be used in bimolecular photochemical processes.
12
Energy
1MLCT
MC(d-d)
kISC
kd-d
∆Eact
3MLCT
hν
ν
knr
kr
∆EES - GS
S0
Figure 2.2 Deactivation paths for the excited state. kisc = rate constant for the intersystem
crossing from the singlet to triplet state, knr = rate constant for non radiative decay from the
3
MLCT states, kr = rate constant for radiative decay from the 3MLCT states, ∆Eact = activation
barrier for population of the MC state, kd-d = rate constant for population of the MC state,
∆EES - GS = energy difference between the ground state and 3MLCT states.
The excited state energy of ruthenium polypyridine complexes can be changed
by using different ligands [20]. If the same metal coordinates different ligands
the lowest MLCT state will be localized on the ligand that is easiest to reduce. It
is also possible to fine tune the excited state energy with different substituents
on the ligands, that can be either electron donating or withdrawing. This will
change the electron density on the ligand and the energy of the MLCT transition
will be affected. Investigations have shown that a linear relation exists between
the difference of the redox potentials, for the oxidation of the metal and
reduction of the ligand, with the energy of the corresponding MLCT transition
[33]. This correlation supports a picture where the orbitals involved in the
redox-processes are the same as those involved in the charge transfer transitions:
EMLCT (eV) = Eox - Ered + K
(2)
where K includes contributions from inner (vibrational) and outer (solvent)
reorganization and the difference in Coulombic interaction between the metal
and the ligand for the redox states and the MLCT state.
13
2.2 Excited state reactions
Due to the higher energy content of the excited state, the excited Ru(bpy)32+ is
both a better reductant and oxidant than the ground state. An additional pathway
for deactivation of the excited state except the intrinsic decay pathways can be
electron or energy transfer to another molecule, where the electron transfer can
be either reductive or oxidative.
Oxidative electron transfer
Reductive electron transfer
Energy transfer
Ru2+* + Q → Ru3+ + Q−
Ru2+* + Q → Ru+ + Q+
Ru2+* + Q → Ru2+ + Q*
(3)
(4)
(5)
Which mechanism that will dominate depends on the properties of the donor and
acceptor. The feasibility of the different deactivation reactions of the excited
state can be estimated from the reaction free energy.
In electron transfer reactions the redox potentials of the donor and acceptor
ground states can together with the zero-zero excitation energy, E0-0, be used to
calculate the redox potentials for the excited state, if the entropy change between
the ground state and excited state is assumed to be negligible [20].
E0 (Ru3+/Ru2+*) = E0 (Ru3+/Ru2+) − E0-0
E0 (Ru2+*/Ru+) = E0 (Ru2+/Ru+) + E0-0
(6)
(7)
Then the standard free energy change, ∆G0, of the reactant and product states,
neglecting the coulombic stabilization energy, can be calculated according to:
∆G0 = E0 (Ru3+/Ru2+*) – E0 (Q+/Q)
∆G0 = E0 (Q+/Q) – E0 (Ru2+*/Ru+)
(8)
(9)
If energy transfer is going to contribute as a deactivation path the energy of the
excited state of the acceptor must be equal or lower than the energy of the donor
excited state. For energy transfer processes the standard free energy change
between the reactant and product states can be calculated from the difference of
the zero-zero spectroscopic energies, E0-0, of the donor and acceptor excited
states according to [34]:
∆G0 = E0-0 (A*/A) − E0-0 (D*/D)
(10)
14
E0-0 can be calculated from the absorption and emission spectra of the donor and
acceptor. If the excited state is a triplet state E0-0 is sometimes estimated from a
fit to the emission spectrum [35].
The excited state can be deactivated by the acceptor, Q, in a bimolecular
reaction where the donor and acceptor are diffusing in the solution. The
quenching rate constant, kq, is then evaluated according to a reaction scheme
which includes the diffusion of the reactants and products. The actual electron or
energy transfer step takes place in an encounter complex where the reactants are
at short separation distance. In most cases a simplified steady state
approximation is valid, giving:
kq = (kd ke) / (k-d + ke)
(11)
where kd and k-d are the rate constants for the diffusion of the reactants and ke is
the electron or energy transfer rate constant within the encounter complex. The
value of the quenching rate constant can be determined from a Stern-Vollmer
relationship. Note however, that the value of ke can never be directly determined
from eq. (11), and that kq can never exceed kd, not even for large ke. An
intramolecular reaction between a donor and an acceptor in a preformed
complex is not limited by diffusion and can be much more rapid. If the donor
and acceptor are connected through bridging units that enhance the electronic
coupling electron, and energy transfer (electron exchange) can occur via the
orbitals of the bridge (“superexchange”) [36].
2.2.1 Photoinduced electron transfer
In quantum mechanical models the golden rule expression for the transition
probability between different electronic states, is often used to treat non
adiabatic electron transfer. In the high temperature limit, when the energy of
each vibration is considerably less than the thermal energy, hν<<kBT, the rate
constant can be written [37,38]:
kET = (2π/h) H2 ( 4πλkBT)−1/2 exp (-∆G#/kBT);
∆G#=(λ+∆G0)2/4λ
(12)
where H is the matrix element for electronic coupling, λ is the reorganization
energy, kB is the Boltzmann constant, and ∆G# is the activation free energy. In
the expression for ∆G#, ∆G0 is the standard free energy change for the reaction.
The rate expression predicts (if the only parameter that is changed is ∆G0) that
the rate of electron transfer will increase with a more negative ∆G0 until the
reaction is activationless, i.e. when -∆G0 is equal to the reorganization energy, λ.
At this point the rate of electron transfer will be at its maximum. When -∆G0>λ
15
the reaction will be in the inverted region and the rate will decrease again. The
reorganization energy λ is the sum of the inner, λin (vibrational, structural
rearrangement within the molecule), and outer, λout (solvent, rearrangement of
the solvent molecules to solvate the new charge situation), reorganization
energies. λout can be estimated from a dielectric continuum model for the
solvent, and give the largest contribution to λ in many electron transfer reactions
in polar media [38]. λin can be calculated with knowledge of the reduced force
constant, for the vibration involved in the electron transfer reaction, and the
equilibrium bond lengths of the reactant and product states [38]. In paper III the
temperature dependence of the rate constant for electron transfer between
ruthenium and manganese was measured in different Ru(II)-Mn(II) complexes.
The resulting value of the total reorganization energy was unusually high which
could indicate that large structural rearrangements (inner reorganization) are
necessary when Mn(II) is oxidized.
2.2.2 Photoinduced energy transfer
Nonradiative electronic energy transfer between a donor and acceptor in a
covalently linked complex can occur if there is some interaction between the
two states. This interaction can be divided into two parts, the Coulombic
interaction (Förster mechanism) [39,40] and the exchange interaction (Dexter
mechanism) [41]. Both mechanisms can act in parallel but due to the nature of
the transitions involved and the distance between donor and acceptor, it is often
possible to determine which mechanism that is the dominant one.
The Förster mechanism is a through space mechanism based on Coulombic
interactions that does not require orbital overlap of donor and acceptor. The
Coulombic mechanism is effective when the involved transitions have high
oscillator strengths, typically for spin allowed processes such as singlet-singlet
energy transfer. This mechanism can occur over distances up to 100 Å. The
Dexter mechanism is a short range mechanism, with a distance between donor
and acceptor typically shorter than 10 Å, that requires that the orbitals of donor
overlap with the orbitals of the acceptor since electrons are exchanged. In
contrast to the Förster mechanism, energy transfer according to the Dexter
mechanism can occur for spin forbidden processes. The selection rules for the
Dexter mechanism requires spin conservation of the reacting pair as a whole.
This makes transitions like D*(T1)A(S0)→D(S0)A*(T1) allowed. The exchange
interaction can be effective over larger distances in linked donor-acceptor
molecules, if the connecting bridge allows electronic interactions via the orbitals
of the bridge (superexchange, in analogy with electron transfer)
16
Coulombic Energy Transfer
EnT
D*
+
A
D
+
A*
+
A*
Exchange Energy Transfer
EnT
D*
+
A
D
Figure 2.3
Förster and Dexter energy transfer. In the Förster mechanism the dipole-dipole
interaction of the donor and acceptor leads to excitation of an electron of the acceptor while
the donor is de-excited. The excitation energy is transferred trough space without a need for
close contact between the reacting species. In the Dexter mechanism the excited electron on
the donor is transferred into the lowest unoccupied orbital of the acceptor with a simultaneous
transfer of an electron from the highest occupied orbital of the acceptor to the corresponding
orbital of the donor.
A large coupling through the bridge will favor the Dexter mechanism (trough
bond) while a more saturated bridge will favor the Förster mechanism (trough
space) [42]. The rate constant according to the Förster mechanism can be
expressed as [39]:
k=
9000(ln 10 )κ 2φ D
128π 5 n 4 N A r 6τ D
∞
J = ∫ FD (ν )ε a (ν )
0
(13)
J
dν
(14)
ν4
where φ0D is the quantum yield of the donor in the absence of acceptor, n is the
refractive index of the solvent, NA is Avogadro´s number, r is the separation
distance between donor and acceptor and τD is the excited state lifetime of the
donor. κ2 is a factor describing the relative orientation of the transition dipoles
for the donor and acceptor respectively. J is the overlap integral that can be
calculated from the normalized emission spectrum of the donor and the
17
absorption spectrum of the acceptor. Thus the rate constant can be calculated
from spectroscopic properties of the donor and acceptor.
According to the Dexter (electron exchange) mechanism the rate constant for
energy transfer can be expressed as [41]:
k = (4π2 / h) Z2 J′
(15)
Where Z is related to the electron matrix element for electron exchange and J′ is
the overlap integral. The parameter Z2 is proportional to exp(-2r/l) where r is the
distance between donor and acceptor and l is the sum of the van der Waals radii
of the donor and acceptor molecules. Thus, the rate is expected to decrease
exponentially with increasing distance between donor and acceptor. The
exchange interaction does not depend on the oscillator strengths of the
transitions involved in contrast to the Coulombic interaction.
In the non-adiabatic limit the energy transfer according to the electron exchange
mechanism can be treated according to the semi-classical expression for electron
transfer reactions [43,44]. In the expression for the rate constant (equation 12),
the reaction free energy ∆G0 and the reorganization energy, λ, can be estimated
from spectroscopic properties of donor and acceptor respectively, ∆G0 from the
difference of the zero-zero spectroscopic energies and λ from the Stoke´s shift
[34]. If ∆G0 and λ have been estimated the electronic coupling between donor
and acceptor can be obtained if the rate constant for the energy transfer process
have been measured. In chapter 4 ultra fast energy transfer between Ru(II) and
Os(II) at close separation distance is described. The general treatments described
in this section are difficult to apply because the ultra fast energy transfer occurs
between states that are not thermally relaxed.
2.3 Time resolved studies of electron and energy transfer
Time resolved emission measurements of the excited state lifetime of Ru(bpy)32+
can give valuable information of the quenching process. For intramolecular
processes the rate constant for electron and energy transfer can be obtained by
measuring the lifetime of the donor with and without acceptor. In an energy
transfer process it can also be possible to measure the rise time of the emission
of the acceptor.
If the donors and acceptors that are involved in energy and electron transfer
reactions have well known absorption spectra in the different electronic states,
transient absorption spectroscopy is a useful technique to monitor the reactions.
18
A
B
Sn
Sn
S1
S1
hν
ν
+A
Photochemical
reaction
S0
S0
Donor*
Donor
Figure 2.4 A: ground state of the donor molecule, D, B: excited state of D.
In the transient absorbance technique the sample is excited with a laser pulse of
a short time duration and then the absorbance is monitored as a function of time.
The length of the laser pulse is crucial for what type of reactions that can be
studied with a certain laser setup. Short femtosecond laser pulses make it
possible to study very fast energy and electron transfer reactions in covalently
linked complexes. For slower reactions, like diffusional quenching or the study
of triplet absorption, the nanosecond flash photolysis technique is more useful
since it provides a longer time window for detection. In this thesis both
techniques have been used. In Figure 2.4 a model system (A) is excited with a
laser pulse of suitable wavelength to create the excited state of D (B). In the
transient absorption spectrum this will be seen as a ground state bleach and
excited state absorption. When a photochemical reaction takes place the excited
state absorption decays, and the absorption of the products formed in the
reaction will appear. If the time window is long enough also the back electron
transfer can be studied. At each monitoring wavelength the transient absorption
signal will be a sum of the ground state bleach, excited state absorption and the
absorption from the species formed in the reaction at different time delays after
excitation. If there will be a net bleach or absorption depends on the exctinction
coefficients for the species involved in the reaction. If stimulated and
spontaneous emission is significant, the contribution to the total signal will
appear as a bleaching. A description of the experimental setup used in the
nanosecond and femtosecond measurements is given in the Appendix.
19
Chapter 3
Mimicking the Electron Transfer Reactions of
Photosystem (II)
In a Swedish collaboration we have synthesized and studied supra molecular
complexes which intend to mimic the donor side of Photosystem II. Several
systems have been made based on manganese (II) and tyrosine covalently linked
to Ru(bpy)32+ [45,46,47,48,49]. The aim was to investigate the possibility of
electron transfer from Mn(II) or tyrosine to photo-oxidized Ru(III) after
quenching of the excited Ru(II)* by an external electron acceptor. More
sophisticated systems have also been made where Ru(II) is linked with a
tyrosine and two Mn (II) ions [50].
In this chapter the first results which was obtained on intramolecular electron
transfer from a covalently linked Mn(II) or tyrosine to photo-oxidized Ru(III)
will be presented (Paper I and II, III]. Other processes that may complicate the
electron transfer as quenching of the excited state of Ru(II)* by the linked
Mn(II) and dissociation of Mn(II) from the Ru(II)−Mn(II) complex will be
discussed. In the last section some preliminary results of electron transfer in a
covalently linked triad, will be presented.
3.1 Electron transfer from Mn(II) linked to photo-oxidized
Ru(III)
3.1.1 Quenching of the excited state of ruthenium by Mn(II)
For efficient electron transfer to an external acceptor the lifetime of the excited
state is of crucial importance. If the lifetime is too short the excited state will
decay to the ground state before any photochemical reaction has time to occur.
The lowest excited state of Ru(bpy)32+ is the 3MLCT state and the lifetime in
acetonitrile at room temperature is ≈900 ns, which is enough for a diffusion
controlled reaction with an acceptor. Without coordinated Mn(II) the lifetime of
the complexes shown in Figure 3.1 is 980, 1050, and ≈900 ns respectively,
which is approximately the same as for Ru(bpy)32+. However when Mn(II) is
coordinated the lifetimes decrease to 250, 7, and ≈300 ns respectively. The
desired reaction for the Ru(II)-Mn(II) complexes is intramolecular electron
20
transfer from the coordinated Mn(II) to photo-oxidized Ru(III) and not a
competing process that quenches the excited state of ruthenium before an
electron is transferred to the external acceptor. Therefore it is important to make
the quenching of the excited state of Ru(II) as slow as possible but at the same
time the electron transfer from Mn(II) to Ru(III) should be fast and efficient.
Cl
H2O
N
Cl
Mn
N
OH2
Cl
N
N
Cl
N
N
Ru
N
N
1
N
2+
N
N
N
Ru
N
N
N
N
N
N
Mn Cl
N
2
N
2+
N
N
Mn Cl
N
Ru
N
N
3
N
2+
N
N
Figure 3.1 Structure of the Ru(II)-Mn(II) complexes. The metal-to-metal distance is 13, 9 and
14 Å respectively, estimated from simple molecular modeling.
Different reasons for the shorter lifetime of the MLCT state of Ru(bpy)32+ can be
considered. If the electronic structure of ruthenium is changed when Mn(II) is
coordinated, the energy of the MLCT state can be changed which will affect the
intrinsic deactivation rate constants. Due to the small variation of the absorption
and emission spectra compared to the complexes without Mn(II) this is not
believed to be the explanation for the shorter lifetimes in the Ru(II)-Mn(II)
complexes. The quenching is therefore attributed to a new deactivation
mechanism in the presence of coordinated Mn(II).
Different mechanisms are conceivable; A. Electron transfer from Mn(II) to
excited Ru(II)* generating Mn(III) and Ru(I). B. Energy transfer to an excited
state of Mn(II) from the excited Ru(II) generating Mn(II)* and Ru(II). C. A third
possibility is paramagnetic quenching which for ruthenium complexes is
believed to be of no significance [51]. To get a better understanding for the
quenching process, low temperature emission lifetime measurements were
performed on 1 (Paper I) as well as on a number of other Ru(II)-Mn(II)
complexes (Paper III). The results indicated that the quenching occurred by
energy transfer and not electron transfer. Approximately the same lifetime of the
intact Ru(II)-Mn(II) complexes was obtained at 77 K as at room temperature,
indicating that the reaction free energy, ∆G0, for electron transfer was not
significantly decreased at lower temperature. For quenching via electron transfer
a decrease of the driving force is expected at low temperature since the solvent
21
cannot reorganize in response to the new charge distribution. Thus, ∆G0 would
increase by an amount equal to the reorganization energy which would be c. a.
1.0 eV [52,53,54]. Since the calculated driving force was close to zero already at
room temperature, electron transfer quenching would be strongly endothermic at
77 K. For energy transfer the outer reorganization is much smaller than for
electron transfer and thus the temperature difference would have a smaller effect
on the quenching rate, consistent with our observations. Förster-type energy
transfer can be ruled out by the very weak absorption of Mn(II)-complexes in
the visible region, since the transitions in Mn(II) are both symmetry and spin
forbidden. However energy transfer via the Dexter mechanism may be spinallowed for energy transfer from the 3MLCT state of Ru(II). The quenching rate
constant was found to decrease exponentially with increasing distance between
Ru(II) and Mn(II) (Paper III) and this is consistent with energy transfer via an
exchange mechanism. The conclusion drawn is that in order avoid a too fast
quenching of the excited state of Ru(II)* the distance between the coordinated
Mn(II) and Ru(II) should be kept long enough. Unfortunately this is probably in
conflict with fast and efficient electron transfer after photo-oxidation of Ru(II).
One possible solution to this problem was to introduce tyrosine as a redox
intermediate between Mn(II) and Ru(II). Emission and transient absorption
measurements showed that tyrosine did not quench the excited state of
ruthenium compared to a model compound, and after oxidation of ruthenium
tyrosine was also capable of transferring an electron to Ru(III) regenerating
Ru(II). In PS II, TyrosineZ is positioned between P680 and the manganese
cluster functioning as a redox intermediate [1b,4]. Maybe this is nature´s
solution to avoid quenching of P680* before the desired electron transfer
reactions occur. In section 3.2 the results on electron transfer studies from
tyrosine to photo-oxidized Ru(III) will be presented.
It has also been shown in paper III that strong electron withdrawing groups on
the Ru(II)-ligands that are remote from the Mn(II)-moiety make the quenching
of the excited state much slower, while the electron transfer rate from Mn(II) to
Ru(III) is even somewhat higher. Since the lowest MLCT state will then be
localized on the ligand easiest to reduce [20], electron withdrawing groups on
the bipyridines results in localization of the electron to those ligands which in
this case will increase the effective distance between the Ru(II) excited state and
Mn(II).
3.1.2. Electron transfer from Mn(II) to photo-oxidized Ru(III)
The first successful intramolecular electron transfer from Mn(II) to Ru(III) was
obtained in complex 1 (Paper I). At this time it was still unknown if
22
intramolecular electron transfer was going to occur. From the redox potentials of
MnII/III (0.9 V) and RuII/III (1.3 V) (vs SCE in CH3CN), the driving force for
electron transfer was calculated to be 0.4 eV. Thus it seemed thermodynamically
feasible to reduce Ru(III) by the coordinated Mn(II). Transient absorption
measurements showed that after the competing Ru(II)* quenching processes by
Mn(II) and the external electron acceptor, MV2+, the charge separated state was
formed creating Ru(III) and the reduced acceptor MV+• in equimolar
concentrations[55,56,57]. The ground state bleach of Ru(II) due to formation of
Ru(III) was monitored at 450 nm. The formation and decay of the reduced
acceptor was monitored at 600 nm where MV+• is known to absorb. The
oxidation of Mn(II) could not be observed in the transient absorption
measurements due to the very weak absorption of both Mn(II) and Mn(III).
Cl
H2O
e-
Cl
Mn
OH2
N
N
2
hν
ν
N
N
Ru
N
N
2+
eN
1
MV2+
N
Figure 3.2 1: Electron transfer from the excited state of Ru(II)* to MV2+. 2: Intramolecular
electron transfer from Mn(II) to Ru(III).
In the model compound the recovery at 450 nm and the decay at 600 nm
occurred with the same second order rate constant, attributed to recombination
between Ru(III) and MV+•. In Ru(II)-Mn(II) the decay of MV+• occurred with
the same second order rate constant as for the model compound without
coordinated Mn(II). However the recovery of Ru(II) observed at 450 nm was
much faster. The ligand was not a possible electron donor on the observed
timescale that was shown in the measurements with the model compound. The
only source present except MV+• capable of giving an electron to oxidized
ruthenium was then Mn(II). Since the electron did not come from MV+• the
conclusion was that it must have come from Mn(II). Thus the desired electron
transfer from Mn(II) to photo-oxidized Ru(III) occurred.
23
The electron transfer between Mn(II) and Ru(III) was also confirmed by EPR
measurements where chemically produced Ru(bpy)33+ was mixed with an
equimolar amount of complex 1, which contains Mn(II). Ru(III) and Mn(II)
show characteristic EPR signals whereas Ru(II) and Mn(III) are EPR silent. The
signals disappeared after mixing which showed that Ru(III) was capable of
oxidizing Mn(II) [46].
However a dependence of the rate constant on the concentration of the Ru(II)Mn(II) complex was observed and the kinetic traces were not simple single
exponentials. This is not expected for an intramolecular electron transfer
reaction. Thanks to the Ru(II)* quenching process, steady state and time
resolved emission measurements could show that Mn(II) dissociated from the
complex leaving one fraction without bound Mn(II) in the solution. These
fractions could be quantified from a biexponential fit of the emission decay. A
kinetic model was assumed were intramolecular electron transfer occurred in the
intact complex, RuIIILMnII, and in the complexes without bound manganese,
RuIIIL, electron transfer could only occur after reassociation of Mn(II). In Figure
3.3 the reaction scheme for the processes included in the kinetic model is shown.
Figure 3.3 Reaction scheme for the electron transfer processes included in the kinetic analysis
of the recovery of Ru(II) at 450 nm.
The disappearance of Ru(III) was observed as a recovery of the bleach at 452
nm after the decay of the excited state. The disappearance of the total
concentration of Ru(III) after the decay of the excited state is given by:
24

−k t
k Mn II   Ru III L  

 
 0
e et
  Ru III LMn II  +  Ru III L  (t ) =   Ru III LMn II  − a 

+

 
 
 0

 Mn II   k  MV +  t + 1
 
k
k
−

et
a 
  rec 
 0

− k [MnII ] t
k  Ru III L 
et 
 0
e a
+
k − k  Mn II  k MV +  t + 1
et
a 
 rec 
 0
(16)
The parameters ket and ka[MnII] and the total concentration of Ru(III) ([RuIII]0tot
= [RuIIILMnII]0 + [RuIIIL]0) initially formed, were determined from a fit of the
kinetic traces to equation 16 at each total concentration of complex 1. The initial
fractions of [RuIIILMnII]0 and [RuIIIL]0 and the value of krec[MV+]0 were fixed in
the fitting procedure and determined in independent experiments. The resulting
values of ket and ka[MnII] from the curve fit were plotted versus the
concentration of free Mn(II). As can be seen in Figure 3.4, ket is constant with
increasing concentration of free Mn(II) whereas ka[MnII] shows the expected
linear concentration dependence. This was assumed in the kinetic model and the
results obtained were consistent with that. The rate constant for intramolecular
electron transfer from Mn(II) to Ru(III) was obtained as ket=1.8 × 105 s-1 and the
rate constant for association of Mn(II) with the dissociated complex was
obtained as 2.9 × 109 M-1s-1.
Control experiments were needed for further support of the kinetic model, that is
not unique. Intermolecular electron transfer from Mn(II) to photogenerated
Ru(III) on different complexes could also be an explanation for the
concentration dependent process. However stopped-flow experiments, in which
chemically produced Ru(III)(bpy)3 was mixed with complex 1 and the
disappearance of Ru(III) was monitored optically, ruled out this explanation.
The obtained rate constant for the bimolecular process was too slow (≈1 × 106
M-1s-1), to explain the observed concentration dependence in Figure 3.4 Electron
transfer from dissociated Mn(II), free in solution was excluded based on
experiments where Ru(bpy)32+ was titrated with MnCl2 and the emission of
Ru(bpy)32+ was measured.
Intramolecular electron transfer also occurred in complexes 2 and 3, with rate
constants ket>2×107 s-1 and ket≈1×105 s-1 respectively. The short lifetime of the
excited state of 2 required a very high concentration of the external acceptor,
MV2+, to compete with intramolecular quenching of the excited state of Ru(II)*
25
Figure 3.4 Rate constant for intra molecular electron transfer, ket versus the concentration of
free Mn(II) (triangles) and psedo-first order rate constant, ka[MnII]free versus free Mn(II)
(circles).
by Mn(II). The dissociation of Mn(II) in complexes 2 and 3 was less significant
than for complex 1 and a simpler kinetic analysis could be applied.
Temperature dependent measurements of the electron transfer rate in different
Ru(II)-Mn(II) complexes (Paper III) have shown that the inner reorganization
energy is unusually high in these complexes. This is probably due to a large
reorganization of the coordination site when Mn(II) is oxidized to Mn(III). This
could indicate that this type of manganese complexes never will be fast electron
donors. This was another reason −except to avoid fast quenching of the excited
state of Ru(II)*− for introducing a redox intermediate such as tyrosine between
ruthenium and manganese. Then tyrosine could act as a fast electron donor to
regenerate Ru(II). A third reason for introducing an intermediate electron donor
is to avoid reduction of higher oxidation states of the manganese complex by the
excited Ru(II)*. This will be important in multistep oxidation of the manganese
complex where the intention is oxidation of water. Our first attempt of electron
transfer from tyrosine to photo-oxidized ruthenium will be presented in the next
section.
3.2 Electron transfer from tyrosine to photo-oxidized Ru(III)
In the last section, three reasons for introduction of an intermediate electron
donor between Ru(II) and Mn(II) was mentioned that are of particular interest
26
for avoiding unwanted reactions of the Ru(II)* excited state. In addition, it also
seemed interesting to use tyrosine as an electron donor to further mimic the
reactions of PS II.
A complex was synthesized where ruthenium was covalently linked to a tyrosine
residue via an amide bridge (Figure 3.5). Time resolved emission measurements
showed that the excited state decay was not affected by the linked tyrosine. The
lifetime in water solution at pH=7 was 370 ns, the same as for the model
compound where ruthenium was linked to an alanine. The redox potential of
tyrosine is +0.97 V in water at pH=7 [47]. This makes electron transfer to
Ru(III) thermodynamically feasible, ∆G0 = −0.3 eV. Transient absorption and
EPR measurements were used to investigate the intramolecular electron transfer
reaction.
-H+
OH
e-
2
COOEt
HN
hν
ν
O
N
N
Ru
N
N
eN
1
MV2+ or [Co(NH3)5Cl]2+
N
Figure 3.5 1: Electron transfer from excited Ru(II)* to MV2+ or Co3+. 2: Intramolecular
electron transfer from tyrosin to Ru(III) with a simultaneous deprotonation [48].
Excitation of the Ru(II)-moiety of the Ru(II)-Tyr complex in the presence of the
external acceptor MV2+ resulted in charge separation. The appearance and decay
of MV+• was monitored at 600 nm and the bleaching and recovery of Ru(II) was
monitored at 450 nm. After the quenching of Ru(II)* by MV2+ was completed,
the decay of the oxidized Ru(III) and reduced MV+• was followed kinetically.
The recovery of Ru(II) occurred with a rate constant of k= 5 × 104 s-1, while the
decay of the absorbance signal from MV+• was much slower. The conclusion
was that intramolecular electron transfer from tyrosine occurred regenerating
Ru(II). For the model compound (without tyrosine) it was found that the decay
27
of MV+• and recovery of Ru(II) occurred with the same second order rate
constant (k=8 × 109 M-1s-1) consistent with a diffusion controlled recombination
between MV+ and Ru(III). Further support that Ru(II) was regenerated by the
tyrosine moiety was obtained when a sacrificial electron acceptor,
Co(NH3)5Cl2+, was used that also give negligible transient absorption changes in
the near UV-vis region upon reduction.
Figure 3.6 Kinetic traces at 450 and 410 nm for Ru-Tyr and Ru-Ala, with Co(NH3)5Cl2+ as
electron acceptor. The increased absorption at 410 nm was attributed to absorption from the
tyrosine radical, formed in the electron transfer reaction with Ru(III). For the model
compound Ru-Ala, recovery of Ru(II) was not observed on the measured time scale.
At 410 nm, where tyrosine radicals has been reported to absorb [58] an
absorption appeared, after the recovery of the initial bleaching. This was
attributed to absorption from the oxidized tyrosine. The recovery of the 450 nm
signal for the Ru(II)-Tyr complex occurred with the same rate as when MV2+
was used, whereas the signal at 450 nm for the model compound did not decay
at all in the observed time window. To rule out the possibility of intermolecular
electron transfer between Ru(III) and tyrosine on different complexes, a control
experiment was performed where a solution of Ru(bpy)32+, MV2+ and free
tyrosine was flashed. However, the rate constant for the bimolecular reaction
was two orders of magnitude too small to account for the rate obtained in the
experiments with the Ru(II)-Tyr complex.
EPR measurements confirmed that a light-induced oxidation of the tyrosine
moiety occurred. Important information was also obtained on the lifetime of the
oxidized tyrosine radical, which could not be followed in the transient
28
absorption measurements at 410 nm. An EPR spectrum was recorded of the
Ru(II)-Tyr complex during illumination in the presence of the sacrificial
acceptor Co(NH3)5Cl2+. An EPR signal centered around g=2.0044 was obtained
and attributed to the oxidized tyrosine residue, due to the high g-value that is
very similar to the g-values of other neutral tyrosyl radicals [59,60] The second
order decay of the radical was measured by time resolved EPR and the first halflife was found to be 50 ms. Intramolecular electron transfer from a phenol type
ligand to a photo-oxidized Ru(II)-complex has also recently been reported by
Wieghardt et al.[61,62].
In a later publication it has been shown that the Ru(II)-Tyr complex was capable
of oxidizing a manganese (III/III) dimer [47]. Other interesting results have also
been obtained concerning the pH-dependence of the electron transfer rate in the
Ru(II)-Tyr complex [48]. The fact that intramolecular electron transfer occurred
in the Ru(II)-Tyr complex, and the following investigations were each a step
further, towards mimicking the reactions in PS II. A complex where Ru(II) is
linked to a Mn(II/II) dimer via a tyrosine residue has been prepared in which
intra molecular electron transfer from the Mn(II/II) dimer to the photo-oxidized
Ru(III) occurred with a rate constant larger than 1 × 107 s-1 (limited by the
quenching of the excited state by the external acceptor) [50]. To be able to
resolve the electron transfer from the manganese dimer a faster acceptor has to
be used. A covalently linked electron acceptor might solve this problem. To
ultimately achieve the goal of water oxidation higher oxidizing states of the
manganese complex have to be used, and work is in progress with modified
ligands to stabilize higher valence states of the Mn-dimer. For a recent review
on the results so far on the Mn2-Tyr-Ru(II) complexes intended to mimic the
donor side of PS II see reference [63].
3.3 Electron transfer in a covalently linked triad
To eventually be able to produce H2 at a catalytic metal center, the development
of artificial models of the acceptor side of PS II is very interesting. In addition
faster electron acceptors are needed to be able to oxidize Ru(II), when the
quenching of the Ru(II) excited state by Mn(II) is too rapid, and to resolve the
intramolecular electron transfer from Mn(II) to photo-oxidized Ru(III).
In order to develop a complex where the electron can be transferred to a second
acceptor (Figure 3.7), the competing back electron transfer from the primary
acceptor to the oxidized photo-sensitizer must not be too fast. Charge separation
in a covalently linked Ru(II)-MV2+ donor acceptor complex and the
corresponding back electron transfer had been reported to occur rapidly, with a
29
e-
hν
ν
1
3
e-
D
e-
e-
2
4
A1
S
e-
2H+
A2
5
e-
H2
Figure 3.7 Schematic picture of a donor-acceptor system for hydrogen evolution. 1:
Excitation of the photosensitizer, 2: electron transfer to the primary acceptor, 3: fast electron
donation to stabilize the charge separated state, 4: electron transfer to a second electron
acceptor which after a second cycle can be double reduced, 5: the electrons are used to
produce H2.
time constant for the back reaction of ≈25 ps [64]. To increase the probability of
avoiding the back reaction, MV2+ was exchanged for quinone since the forward
and backward electron transfer steps had been reported to be slower [65].
Another reason for using quinone as electron acceptor is their capability to
accept two electrons, which is interesting for the reduction of 2H+ to H2.
2+
O
N
N
N
Ru
N
N
N
N
O
N
5+
O
N
N
N
Ru
N
N
N
N
O
N
Co N
N N N
Figure 3.8 Ru2+−Q (left) and Ru2+−Q−Co3+ (right).
Two complexes were made Ru2+−Q and Ru2+−Q−Co3+, where a second electron
transfer step from Q−• to Co3+ was expected to be thermodynamically feasible
from redox data [66]. Femtosecond transient absorption measurements showed
that, the forward and backward electron transfer reactions in the Ru2+-Q
complex occurred with a ≈200 ps and ≈900 ps time constant, respectively.
30
However in the Ru2+−Q−Co3+ complex the back electron transfer from Q−• to
Ru(III) (τ≈900 ps) dominated. To solve this problem and be able to eventually
succeed in a second electron transfer step, another compound was synthesized
where the electron donor phenothiazine (PTZ) was covalently linked to the
Ru2+-Q complex (Figure 3.9). PTZ has been reported to be a fast electron donor
to Ru(III) in the covalently linked triads PTZ-Ru2+-AQ (phenothiazineruthenium-anthraquinone) [67], PTZ-Ru2+-MV2+ [68] and in PTZ-Ru2+-DQ2+
[69]. If PTZ could work as a fast electron donor to Ru(III) in our system the
charge separated state would be stabilized and the back reaction from quinone to
Ru(III) prevented. This would increase the possibility of a second electron
transfer step to a second acceptor. The corresponding dyad PTZ-Ru2+ was also
synthesized as a model compound.
e1
hν
ν
e2
2+
O
N
N Ru N
N
N
N
N
O
N
N
S
3
Figure 3.9 The PTZ-Ru2+-Q triad with electron transfer reactions. 1: Electron transfer from
excited Ru(II) to quinone, 2: fast electron donation from PTZ to Ru(III). 3: Back electron
transfer from reduced quinone to oxidized phenothiazine.
Preliminary results from femtosecond transient absorption measurements
showed that the charge separated state, PTZ+•−Ru2+−Q−• was formed with a time
constant of ≈200 ps, that is approximately the same as the electron transfer
lifetime in Ru2+−Q. This suggests that the initial electron transfer from Ru(II)* is
rate limiting (τ≈200 ps), and that the subsequent PTZ-to-Ru(III) electron
transfer is much faster. An absorption appeared at 510 nm, where PTZ+• is
known to absorb [67], with the same rate as the decay at 450 nm. Measurements
of the model compound PTZ-Ru(II) showed that the excited Ru2+* was
quenched only slowly by PTZ, giving an emission lifetime of 90 ns in degassed
acetonitrile.
A full spectrum was taken with nanosecond flash photolysis to identify the
spectral features of the charge separated state and then measure the rate of the
31
back reaction. The reduced quinone was not observed in the femtosecond
transient absorption measurements, probably due the presence of Ru2+*, of
which the bleaching cancelled the Q−• absorption. However, with the help of a
control experiment where Ru(bpy)32+ was mixed with bensoquinone and a
spectra of the charge separated state was taken, absorption bands corresponding
to the reduced quinone could be identified in the flash photolysis spectrum of
the triad. The back electron transfer:
PTZ+•−Ru2+−Q−• → PTZ−Ru2+−Q
(17)
was found to occur with a time constant of 90 ns, approximately the same as for
the back electron transfer reported for PTZ-Ru2+-AQ [67].
In the femtosecond experiments with PTZ-Ru2+-Q the results indicated that a
fraction of Ru2+* was not involved in the electron transfer to Q. This was further
confirmed in nanosecond emission measurements where a rather large emission
signal was observed at 600 nm, where Ru(II) have emission. This could be
attributed to photo decomposition where the quinone becomes inactive.
Work is under progress where Co(bpy)33+ is being attached to quinone, to create
PTZ−Ru2+−Q−Co, in which the electron hopefully is transferred the second step
to Co(bpy)33+. The Co(bpy)33+ may then be substituted for potentially H2generating catalytic metal complexes.
32
Chapter 4
Mimicking the Light Harvesting Process – Artificial
Antennas
An artificial antenna intended to mimic the light harvesting process in natural
photosynthesis is usually constructed of several molecular components. Each
component has its own special function in the energy transfer process, like
absorption at different wavelengths, lowest excited state energy (energy trap), or
just a bridge connecting two light absorbing components. If the excitation
energy is going to be transferred between many components the rate of each
energy transfer step will be very important. Efficient energy transfer over long
distances, without large losses along the way, is important in molecular devices
such as larger antenna systems [17] and molecular photonic wires [70].
In the first section in this chapter the spectroscopic properties of polynuclear
transition metal complexes are presented, to give a background for the
interpretation of the results obtained in paper V. Thereafter a summary of the
results on ultra fast energy transfer between Ru(II) and Os(II) in a small
artificial antenna fragment will be given, and this will be followed by some
comments of the energy transfer mechanism.
4.1 Artificial antennas based on transition metal complexes
Professor Balzanis group in Bologna and professor Campagnas group in
Messina have synthesized and studied dendrimers of different size, based on
transition metal complexes [17]. The synthetic procedure allows a control of the
exact position of each metal complex. The dendrimers are very interesting for
several reasons. The ability to absorb light increases with the number of metal
centers and multielectron reduction and oxidation can occur at certain potentials.
The dendrimers also show efficient excitation energy transfer to the lowest
excited state. To construct these large supramolecular complexes ruthenium and
osmium have been used as metal centers, 2,3-dpp and 2,5-dpp as bridging
ligands (dpp=bis(2-pyridyl)pyrazine), and bpy and biq (biq=2,2,´-biquinoline) as
terminal ligands.
33
Ru2+
Metal centers
Os2+
N
N
Bridging ligands
N
N
2,3-dpp
Terminal ligands
N
N
bpy
N
N
N
N
2,5-dpp
N
N
biq
Figure 4.1 Structure of the bridging and terminal ligands.
The absorption spectrum of the polynuclear complexes show that the extinction
coefficient increases with the number of metal centers that is incorporated in the
structure. The absorption in the both the UV and visible spectral range can
therefore be very large, with extinction coefficients of 1×105 M-1cm-1 for a
complex with 22 metal centers [18,71]. Several overlapping absorption bands,
corresponding to different MLCT transitions are contributing to the overall
absorption. The energy of the transition depends on the metal and the ligand
involved in the transition, and will largely be determined by the oxidation
potential of the metal and the reduction potential of the ligand. The lowest
excited state will be localized on the ligand easiest to reduce, i. e. with the best
π-accepting ability. The energy of the lowest excited state of an osmium
complex is lower than for the corresponding ruthenium complex. The energy
ordering of the MLCT transitions involving Ru(II) and Os(II) as metal centers
and the ligands in Figure 4.1 is [19]:
Os → µ-2,5-dpp < Os → µ-2,3-dpp < Os → biq < Ru → 2,5-dpp < Ru → 2,3dpp < Os → bpy < Ru → bpy
The first reduction and first oxidation is ligand centered and metal centered
respectively and follows the same pattern as the MLCT transitions. The 2,5-dpp
ligand is easiest to reduce while bpy is most difficult to reduce, and Os(II) is
oxidized at less positive potentials than Ru(II).
34
From redox data it is shown that the interaction between nearby units is
noticeable for metals coordinated to the same ligand and for ligands coordinated
to the same metal but direct interaction between units further apart is not
detectable [72]. Therefore, neighboring ligands coordinated to the same metal
but not directly involved in the transition also affects the energy of the MLCT
transition through their electron donor ability. This will for example make the
energy for Ru(2,3-dpp)32+ higher than for Ru(2,3-dpp)(bpy)22+. The electron
donor ability increases in the series [72]:
µ-2,5-dpp ≤ µ-2,3-dpp < biq < bpy
By varying the metal and the ligands it has been possible to construct complexes
where efficient energy transfer occurs to the lowest excited state in
pre-determined patterns. This has been obtained for complexes with four [73],
six [74], ten [19], and 22 [71], metal centers. Energy transfer from the periphery
to the center and from the center to the periphery has been obtained in some of
the complexes. In larger structures a unidirectional energy transfer will require
an increased number of metals and ligands to accomplish enough variation of
excited state energies. Electron donating or accepting substituents on the
peripheral ligands can also be used to alter the energy of the excited states in the
periphery and thereby creating a gradient for energy transfer.
Efficient energy transfer to the metal unit with the lowest excited state energy
has earlier been demonstrated by steady state emission measurements. It has
been shown that irrespective of excitation wavelength the emission originates
from the component with lowest excited state energy [75]. The assignment of
the emission to certain components is based on comparisons with model
compounds. For example in a binuclear complex
[(bpy)2RuII(2,3dpp)OsII(bpy)2]4+, the ruthenium based emission is expected to occur at a
wavelength comparable to the emission from ruthenium in the model compound
[(bpy)2RuII(2,3-dpp)RuII(bpy)2]4+, and the osmium based emission is expected to
occur at wavelengths comparable to the emission from [(bpy)2OsII(2,3dpp)OsII(bpy)2]4+. In the heteronuclear compound only emission at wavelengths
corresponding to the osmium based emission was detected, suggesting that the
energy transfer from ruthenium to osmium was much faster than the
spontaneous decay to the ground state (τ≈100 ns for the Ru(II)-complexes).
To investigate the rate of energy transfer between two individual metal centers,
two dinuclear Ru-Os complexes and the corresponding homometallic model
complexes were selected. Except of being of fundamental interest the actual rate
35
of energy transfer between the individual units in a large complex intended to
mimic the light harvesting process is of interest for the overall efficiency.
4.2 Ultra fast energy transfer between Ru(II) and Os(II) in a small
artificial antenna
To study the energy transfer process in the Ru(II)-Os(II) complexes (Figure 4.2)
transient absorption spectroscopy was chosen. The reason for this is the low
radiative rate constant for these complexes. The ground state absorption spectra
of the complexes are shown in Figure 4.3 and 4.4. The absorption bands
corresponding to the MLCT transitions in the complexes appear at different
energies in the spectrum.
M2
N
N
N
N
N
N
N
N
N
N
N
N
M1
N
N
N
M2
N
N
N
M1
N
N
N
N
N
N
Ru^Ru
M1=Ru, M2=Ru
Ru–Ru
M1=Ru, M2=Ru, NN=bpy
Ru^Os
M1=Ru, M2=Os
Ru–Os
M1=Ru, M2=Os, NN=biq
Os^Os
M1=Os, M2=Os
Figure 4.2 Structure of the binuclear complexes. Metal complexes with 2,3-dpp as bridging
ligand (right) and metal complexes with 2,5-dpp as bridging ligand (left).
The difference between the absorption spectra of the complexes with the same
bridging and peripheral ligands will depend on the metal centers. A redshift is
observed when Ru(II) is replaced by Os(II). Another interesting feature in the
absorption spectrum is the absorption extending to the red part for the osmium
containing compounds. Due to the enhanced mixing of the singlet and triplet
states by osmium, the probability of absorption from the singlet ground state to
the lowest excited triplet states is slightly allowed and visible in the
36
1,0
Abs.
0,8
0,6
0,4
Os-Os
Ru-Os
Ru-Ru
0,2
A
0,0
400
B
C
D
600
800
1000
Wavelength(nm)
Figure 4.3 Absorption spectrum for the binuclear complexes with 2,3-dpp as bridging ligand,
Ru∧
∧Ru (black), Ru∧
∧Os (dark gray) and Os∧
∧Os (light gray). The different MLCT transitions
indicated with arrows. RU→by and Os→by MLCT transitions at 425 NM (A), RU→2,3-dpp
MLCT at 530 NM (B), Os→2,3-dpp MLCT at 556 NM (C) and the absorption to the 3MLCT
states of Os at 770 NM (D).
absorption spectrum. This allows direct observation of a bleaching that is unique
for the osmium excited state at wavelengths where Ru(II) has no absorption in
the ground state.
In the ruthenium-osmium complexes, the lowest excited state are the triplet
states of osmium, and from the results obtained earlier in the steady state
emission measurements the excitation energy is expected to be transferred to
those states irrespective of excitation wavelength. Because of the overlapping
absorption bands in the spectrum, selective excitation of one metal center is
impossible, but even a 50 % selection of the MLCT transitions of ruthenium
should be observable in the measurements when energy transfer occur with
100% efficiency to the osmium moiety. This would results in a shift of the
excited state spectrum characteristic of ruthenium and osmium to a spectrum
characteristic of only osmium.
The above expectations require that the excited state can be considered as
localized to one metal center at a time. The distance between the metal centers
37
1,2
1,0
Abs.
0,8
0,6
0,4
Ru-Os
Ru-Ru
0,2
A
0,0
400
B
C
D
600
800
1000
Wavelength(nm)
Figure 4.4 Absorption spectrum for the binuclear complexes with 2,5-dpp as bridging ligand,
Ru-Ru (black) and Ru-Os (dark gray). The different MLCT transitions indicated with arrows.
Ru→bpy at 430 nm (A), Ru→2,5-dpp at 580 nm (B), Os→2,5-dpp/biq at 625 nm (C) and the
absorption to the 3MLCT states of Os at 770 nm (D).
in these compounds is approximately 7-8 Å and the coupling is strong enough to
allow very fast energy transfer to the metal unit with the lowest excited state
energy. However the excited state is believed to be localized to one metal center
at a time based on the following arguments.
The luminescence data has shown that the emission from the Ru(II)-Os(II)
complexes agrees well with the emission from the lowest 3MLCT states of
osmium (based on studies on model compounds, Ru-Ru and Os-Os). The
difference of the excited state energies is relatively large 0.2-0.4 eV, and the
redox data shows that oxidation of the metal centers occur at different potentials
characteristic of ruthenium and osmium respectively (1.25 V and 1.55 V) [76].
In the transient absorption measurements the bleaching of the Ru→bpy band is
missing when energy transfer is completed. If the excited state was delocalized
over both metal centers this band should be bleached.
Further support was obtained in an experiment where Ru-Os, was titrated with
the oxidant Ce(IV) and the absorption spectrum measured at each added
concentration. When Ce(IV) was added to oxidize osmium a NIR band
38
attributed to Ru(II)→Os(III) MMCT appeared. It was also initially a decrease in
the intensity of the bands attributed to transitions to the lowest triplet and singlet
MLCT states of the Os-moiety, at ≈800 and 625 nm, respectively. In contrast,
the band at 575 nm attributed to the Ru→2,5-dpp transition (in analogy with the
model complex) was not affected. This shows that the transitions observed are
localized on one or the other metal unit. When more Ce(IV) was added, so that
also the Ru(II) was oxidized, the Ru(II) bands disappeared, as well as the
MMCT NIR band.
(bpy)2RuII(2,3-dpp)OsII(bpy)2
0,01
0,00
-0,01
∆A
-0,02
-0,03
100 ps
3 ps
-0,04
-0,05
-0,06
450
500
550
600
650
700
750
Wavelength(nm)
Figure 4.5 Transient absorption spectrum for Ru∧
∧Os after excitation at 520 nm. The
maximum of the bleaching agrees well with the maximum of the absorption band
corresponding to the Os→2,3-dpp MLCT transition (556 nm).
After excitation of Ru∧
∧Os (Figure 4.5) and Ru-Os (Figure 4.6) on the blue side
of the lowest energy bands, the transient absorption spectra was for both
complexes characteristic of the Os-moiety already after 3 and 5 ps respectively.
No further shift of the spectrum was seen at longer time scales, just decay of the
excited state of osmium (τ≈1 ns). For Ru∧
∧Os the maximum of the bleaching
agreed well with the maximum of the absorption band corresponding to the
Os→2,3-dpp MLCT transition (556 nm) in Os∧
∧Os (Figure 4.3). For Ru-Os the
maximum of the bleaching was at 625 nm, which is the maximum of the
overlapping Os→2,5-dpp and Os→biq transitions. In addition, the bleaching of
the Ru→bpy transition at 450 nm could not be observed.
39
(bpy)2RuII(2,5-dpp)OsII(biq)2
0,05
0,04
0,03
∆A
0,02
0,01
0,00
-0,01
5 ps
-0,02
-0,03
450
500
550
600
650
700
750
Wavelength(nm)
Figure 4.6 Transient absorption spectrum for Ru-Os after excitation at 528 nm. The bleaching
at 625 nm agrees well with the absorption band corresponding to the Os→2,5-dpp/biq
transition. The bleaching of the Ru→bpy band is not observed at 450 nm instead an
absorption have appeared that probably can be assigned to transitions of the reduced 2,5-dpp
[77] or biq [78] of the lowest Os-based MLCT state.
To investigate a possible spectral shift on shorter time scale. Kinetic traces were
∧Os and Ru-Os
recorded at different wavelengths in the spectrum. Both for Ru∧
the results indicated that the energy transfer process occurred faster than the
time resolution, τ<200 fs. A spectral shift was obtained for Ru∧
∧Os with a time
∧Ru
constant of ≈400 fs, however kinetic traces for the model complexes Ru∧
and Os∧
∧Os, showed the same behavior where energy transfer was not expected
to give a spectral shift. The observed spectral shift in Ru∧
∧Os was therefore not
assigned to energy transfer between the metal centers. For further support of
these findings, the polarization dependence of the pump-probe kinetic traces was
investigated for Ru∧
∧Os. With the bent 2,3-dpp as bridging ligand, energy
transfer from one metal center to the other would change the direction of the
transition dipole for the lowest MLCT state, that is believed to be directed along
the metal-ligand axis [24]. This would result in a difference in the kinetic traces
when the probe light is perpendicular or parallel to excitation. However the
kinetic traces did not show a polarization dependence on a < 10 ps timescale,
giving further evidence that the energy transfer in Ru∧
∧Os was faster than the
time resolution. The observed spectral shift with τ≈400 fs was instead attributed
40
to solvent and vibrational relaxation. A similar relaxation time has been reported
for the mononuclear Ru(bpy)32+ [27]. On longer timescale a slower polarisation
decay was observed, attributed to molecular rotation.
The kinetic traces for Ru-Os indicated that the transient spectral features were
characteristic of the excited osmium-moiety at all wavelengths also on a
timescale <5 ps. No changes with time were detected except at 770 nm where a
small increase of the bleaching was observed, with a time constant of τ≈800 fs.
A bleaching of the ground state absorption of the triplet MLCT states of osmium
is expected at 700-800 nm after energy transfer to the osmium-moiety, since the
model complex Ru-Ru has no ground state absorption at these wavelengths.
However the amplitude of the ≈800 fs component is too small to account for
EnT from an initial <50% population of Ru-based excited states. Moreover no
spectral changes were found at any other wavelength, not even at the maximum
of the Os→2,5-dpp/biq bleach (625 nm).
The conclusion was that also in the Ru-Os complex energy transfer occurred
faster than the time resolution. Instead we attribute the observed dynamics at
770 nm (τ≈800 fs), to vibrational and solvent relaxation also for this complex.
When excitation was performed with 800 nm instead, there was no 800 fs
component in the bleach, presumably because the excited state was populated
close to the minimum of the potential energy surface.
4.3 Comments on a possible energy transfer mechanism
Energy transfer in Ru(II)-Os(II) complexes is usually assumed to occur via the
Dexter mechanism due to the spin forbidden nature of the transitions involved
[42]. However the Förster mechanism cannot be excluded due to the partial
singlet character of the lowest excited states of ruthenium and osmium caused
by the enhanced spin orbit coupling. Both mechanisms can act in parallel, but
the relative contribution by the Förster mechanism might be low in rutheniumosmium complexes due to the small radiative rate constant and low value of the
overlap integral of the donor emission and acceptor absorption spectrum. The
rate constant according to the Förster theory can be calculated from
spectroscopic properties [39] and compared with the experimentally obtained
rate constant. This is usually used as a way to estimate the relative contribution
by the Förster mechanism. For the binuclear Ru(II)-Os(II) complexes
investigated in paper V a Förster distance of 2 Å would be required to account
for the rapid energy transfer process (based on comparison with similar Ru-Os
complexes) and since the MLCT transitions are not very strong for these
complexes, not even for the singlet-singlet transitions energy transfer via the
41
Förster mechanism seems unreasonable. If the Dexter mechanism is believed to
be the dominant one, this is described as a double electron exchange and the rate
constant can in the non adiabatic limit be expressed according to the expression
for electron transfer in eq. 12 as described in chapter 2. Energy transfer from
ruthenium to osmium occurred faster than the time resolution, τ<200 fs. It seems
therefore that the excited state dynamics of ruthenium [27,30] and the energy
transfer between ruthenium and osmium in the investigated complexes occur on
comparable time scales. The usual treatment applied for exchange energy
transfer is valid for thermally relaxed states, and this is no longer true for the
ultra fast energy transfer obtained in the present investigation.
When the excitation energy is going to be transferred several steps, in photonic
wires or large antenna structures, the rate of each energy transfer step compared
to the intrinsic decay rate of each metal unit will be important for the overall
efficiency. In absence of osmium, the lifetime of the ruthenium-unit in Ru∧
∧Os
and Ru-Os is ≈100 ns, and if the energy transfer occurs faster than 200 fs, this
means that 99.9998 % of the energy is transferred to the osmium-unit. If the
energy transfer is assumed to occur with the same efficiency in each step the
excitation energy can be transferred over long distances before any substantial
losses of excitation. The investigated dinuclear ruthenium-osmium complexes
are therefore very useful for construction of large and efficient artificial
antennas. The remaining excitation energy can then eventually be used in a
charge separation process in an artificial reaction center.
42
Appendix
Experimental
A.1 Femtosecond pump-probe measurements
The laser system used in the femtosecond transient absorption measurements is
schematically shown in Figure A.1. It consists of two sub-systems that can
operate at the same time, one system (Coherent) that generates femtosecond
laser pulses at a frequency of 200 kHz and one system (Quantronix) that
generates femtosecond laser pulses at a frequency of 1 kHz. For the experiments
performed in this thesis, only the system operating at 1 kHz have been used.
200 kHz
Tunable output
Ar ion laser
MIRA
Nd-YLF
Amplifier, REGA
800 nm
800 nm
OPA
532 nm
Amplifier
800 nm
Strecher/Compressor
800 nm
TOPAS
Tunable output
500-700 nm
1 kHz
Figure A.1 Experimental setup.
The Ar-ion CW laser is pumping the mode-locked Ti-saphire laser (MIRA)
generating laser pulses of 80 fs duration at about 800 nm (bandwidth ≈11 nm).
The 800 nm laser pulses are then amplified by either the 200 kHz system or the
1 kHz system. In the 1 kHz system, a Q-switched Nd-YLF laser generating
frequency doubled nanosecond laser pulses is pumping the Ti-sapphire amplifier
laser. The mode-locked pulses from the Ti-sapphire laser (MIRA) are stretched
in the Stretcher/Compressor and after stretching, these pulses are used to seed
the amplifier. After amplification the output from the amplifier is compressed
and characteristically, the pulses have an autocorrelation halfwidth of 190-200 fs
and pulse energy of about 700 µJ
In the pump-probe measurements with the present setup (Figure A.1) the laser
beam is split into two parts by a beam splitter. 70 % is passed trough an optical
parametric amplifier (TOPAS) which generates visible light tunable between
43
500 and 700 nm. The output from the TOPAS is used for excitation i. e. to pump
the sample. The pump light is passed through a chopper operating at a frequency
between 75-110 Hz before it is focused and overlapped with the probe light in
the sample.
Optical delay line
probe
pump
chopper
sample
Figure A.2 Schematic picture of a pump probe experiment.
The remaining 30 % of the 800 nm is used for probing. To generate white light
the beam is focussed in a sapphire window or CaF2 (in some measurements
water), the white light is then passed through an optical delay line before it is
focused and overlapped with the pump light in the sample cell (1×10 mm). In
the kinetic measurements the detection wavelengths are selected by a
monochromator and then detected by a photomultiplier.
The timescale in the pump probe measurements is obtained by varying the time
delay between the pump and probe. This is accomplished by moving the mirror
of the optical delay line in small steps. In the present setup we have the ability to
use two different delay lines, one that gives a total time window of 1 ns for
detection and one that gives a 5 ns time window. In paper V only the short delay
line have been used, whereas a time window of 5 ns have been necessary for
some of the traces in paper IV. The alignment of the delay line is of crucial
importance in a pump probe experiment. If the beam is not parallel when
entering and exiting the reflecting mirror the overlap with the pump will change
when the mirror is moved. To minimize this effect, the delay line is aligned by
observing the beam at far field. When using the long delay line, kinetic traces of
the bleaching Ru(bpy)32+ was usually measured on a long timescale (50-5000 ps)
as a reference. The lowest 3MLCT state of Ru(bpy)32+ is not expected to decay
significantly on the 5 ns timescale, and the observed trace should not show any
changes. On much shorter timescales, as for the investigation of the energy
44
transfer process in the ruthenium-osmium complexes, only the solvent
(acetonitrile) was measured to exclude that the observed dynamics originated
from intrinsic effects of the solvent.
In the polarization dependent measurements, a polarizer was positioned in the
pump beam and in the probe beam respectively. A λ1/2 plate was positioned
before the white light generation in the probe beam to rotate the fundamental
laser light 50 °. The polarizer in the probe beam was adjusted to select either the
direction parallel or perpendicular to excitation.
The detection of the absorbance change upon excitation is performed with the
Lock-In Amplifier technique. Variations of the probe light which have the same
frequency as the chopper are registrated and the difference between the white
light intensity with and without excitation (I−I0), is obtained. Since the signal
changes are small to the changes in absorption, i.e: (I−I0)/I0 <<1, they are
proportional. A disadvantage is however that the absolute value of the delta
absorption cannot be given since ∆A = −lg (I/I0).
In the spectral measurements a spectrograph (MS 257, Oriel instruments)
equipped with a CCD-detector was used. Instead of a chopper a shutter is placed
in the pump beam, the intensity of the white light is measured with and without
excitation and ∆A is obtained. If spectra are taken at short time delays,
corrections have to be made for the group velocity dispersion. In the
femtosecond measurements presented in this thesis this have not been done
since only kinetic traces have been used to interpret the results on shorter time
scales.
A.2 Nanosecond flash photolysis measurements
In the flash photolysis experiments mainly two different laser systems have been
used. A commercial system from Applied Photophysics, where a Nd-YAG laser
(Spectron Laser system) generating laser pulses of 7 ns duration at 532 nm was
used for excitation. The analysing light was provided from a pulsed xenon lamp,
and the different detection wavelengths were selected with a monochromator
connected to a R 928 photomultiplier. The signal from the photomultiplier was
transferred to a digital oscilloscope.
In the other system, an ELI-94 excimer laser operating with XeCl at 308 nm was
used to pump a dye laser (LT-1113). The output from the dye laser used for
excitation, gave laser pulses at 460 nm with 20 ns duration. The analysing light
was provided by a pulsed xenon lamp and for detection a photomultiplier
coupled to a digitizer (Techtronix) was used.
45
Acknowledgements
First of all I would like to thank my supervisor Leif Hammarström, for your
knowledge, patience and your ability to inspire when everything seems hopeless.
Mats Almgren my supervisor during the first year for support and for creating
such a nice and friendly working atmosphere.
Jan Davidsson for teaching me how to handle the femtosecond laser system and
for always finding time to help.
All present and previous members of The Consortium for Artificial
Photosynthesis are thanked for nice cooperation.
All colleagues and friends at the Department of Physical Chemistry especially
Margit, Laila, Gösta, and Sven for always being very helpful.
Malin för trevligt samarbete, roliga konferenser, och för ditt glada och trevliga
sätt.
Mikael; Vi har slitit tillsammans och under tiden har våra diskussioner och
roliga upptåg i lasergrottan förgyllt tillvaron.
Maria för att du är en god vän. Våra långa telefonsamtal, många gympapass och
trevliga luncher har varit ett stöd och en inspirationskälla.
Min familj
Mormor Judith och farmor Karin samt Allan, Gunnar och Elisabet för att ni
alltid har brytt er om och funnits där.
Mamma och Tomas för stöd och uppmuntran.
Micael, Jonas och Linn: Det som verkligen gör mig lycklig är att se era glada
ansikten.
46
Bibliography
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
a) Deisenhofer, J.; Norris, J. R. In the Photosynthetic Reaction Center, Academic
press: San Diego, 1993, Vol 1-2. b) Oxygenic Photosynthesis: The Light Reactions, ed
D. R. Ort and C. F. Yocum, Kluwer Academic Publishers, Dordrecht, The
Netherlands, 1996.
Sundström, V.; Pullerits, T.; van Grondelle, R. J. Phys. Chem. B 1999, 103, 2327.
a) Yachandra, V. K.; Sauer, K.; Klein, M. P. Chem. Rev. 1996, 96, 2927. b)
Ruettinger, W.; Dismukes, G. C. Chem Rev. 1997, 97, 1.
Hoganson, C. W.; Babcock, G. T. Science 1997, 277, 1953.
Tommos, C.; Babcock, G. T. Acc. Chem. Res. 1998, 31, 18.
Limburg, J.; Vrettos, J. S.; Liable.Sands, L. M.; Rheingold, A. L.; Crabtree, R. H.;
Brudwig, G. W. Science, 1999, 283, 1524.
Wasielewski, M. R. Chem. Rev. 1992, 92, 435.
Sauvage, J.-P.; Collin, J.-P.; Chambron, J.-C.; Guillerez, S.; Coudret, C.; Balzani, V.;
Barigelletti, F.; De Cola, L.; Flamigni, L. Chem. Rev. 1994, 94, 993.
Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 1993, 26, 198.
Kurreck, H.;Huber, M. Angew. Chem. Int. Ed. Engl. 1995, 34, 849.
Imahori, H.; Sakata, Y. Eur. J. Org. Chem. 1999, 2445.
Molnar, S.; Nallas, G.; Bridgewater, J.; Brewer, K. J. Am. Chem. Soc. 1994, 116,
5206.
a) Steinberg-Yfrach, G.; Liddell, P. A.;Hung, S.-C.; Moore, A. L.; Gust, D.; Moore, T.
A. Nature, 1997, 385, 239-241. b) Steinberg-Yfrach, G.;Rigaud, J.-L.; Durantini, E.
N.; Moore, A. L.; Gust, D.; Moore, T. A. Nature, 1998, 392, 479.
Hsiao, J.-S.; Krueger, B. P.; Wagner, R. W.; Johnson, T. E.; Delaney, J. K.; Mauzerall,
D. C.; Fleming, G. F.; Lindsey, J. S.; Bocian, D. F.; Donohoe, R. J. J. Am. Chem. Soc.
1996, 118, 11181-11193.
Lin, V. S.-Y.; DiMagnio, S. G.; Therien, M. J.; Science, 1994,1105.
Kuciauskas, D.; Liddell, P.A.; Lin, S.; Johnson; T. E.; Weghorn, S. J.; Lindsey, J. S.;
Moore, A. L.; Moore, T. A.; Gust, D. J. Am. Chem. Soc. 1999, 121, 8604.
Balzani, V.; Campagna, S.; Denti, G.; Juris, A.; Serroni, S.; Venturi, M. Acc. Chem.
Res. 1998, 31, 26-34 (and references therein).
Campagna, S.; Denti, G.; Serroni, S.; Juris, A.; Venturi, M.; Ricevuto, V.; Balzani, V.
Chem. Eur. J. 1995, 1, 211.
Denti, G.; Campagna, S.; Serroni, S.; Ciano, M.; Balzani, V.; J. Am. Chem. Soc. 1992,
114, 2944-2950.
47
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; von Zelewsky, A.
Coord. Chem. Rev. 1988, 84, 85.
Kalyanasundaram, K. In Photochemistry of Polypyridine and Porphyrin Complexes;
Academis Press: London, 1992.
De Armond , M. K.; Myrick, M. L. Acc. Chem. Res. 1989, 22, 364.
Meyer, T. J. Pure and Appl. Chem. 1986, 58, 1193.
Myrick, M. L.; Blakley, R. L.; DeArmond, M. K.; Arthur, M. L. J. Am. Chem. Soc.
1988, 110, 1325.
Cooley, L. F.; Bergquist, P.; Kelley, D. F. J. Am. Chem. Soc. 1990, 112, 2612.
Malone, R. A.; Kelley, D. F. J. Chem. Phys. 1991, 95, 8970.
Yeh, A. T.; Shank, C. V.; McCusker, J. K. Science, 2000, 289, 935.
Hager, G. D.; Crosby, G. A.; J. Am. Chem. Soc. 1975, 97, 7031
Kober, E. M.; Meyer, T. J. Inorg. Chem. 1984, 23, 3877.
Damrauer, N. H.; Cerullo, G.; Yeh, A.; Boussie, T. R.; Shank, C. V.; McCusker, J. K.
Science, 1997, 275, 54.
Demas, J. N.; Crosby, G. A. J. Am. Chem. Soc. 1971, 93, 2841.
Caspar, J. V.; Kober, E. M.; Sullivan, B. P. Meyer, T. J. J. Am. Chem. Soc. 1982, 104,
630.
Gorelsky, S. I.; Dodsworth, E. S.; Lever, A. B. P.; Vlcek, A. A. Coord. Chem. Rev.
1998, 174, 469, and references therein.
Sutin, N. Acc. Chem. Res. 1982, 15, 275-282.8[
Murtaza, Z.; Graff, D. K.; Zipp, A. P.; Worl, L. A.; Jones; W. E.; Bates, W. D.;
Meyer, T. J. J. Phys. Chem. 1994, 98, 10504.
McConnell, H. M. J. Chem. Phys. 1961, 35, 508.
Jortner, J. J. Chem. Phys. 1976, 64, 4860.
Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265.
Förster, Th. H. Discuss. Faraday Soc. 1959, 27, 7.
Van Der Meer, B. W.; Coker, G.; Simon Chen; S.-Y. In Resonance Energy Transfer,
VCH Publishers Inc., New York, 1994.
Dexter, D. L. J. Chem. Phys. 1953, 21, 836.
De Cola, L.; Belser, P. Coord. Chem. Rev. 1998, 177, 301-346.
Balzani, V.; Bolletta, F.; Scandola, F. J. Am. Chem. Soc. 1980, 102, 2152.
Scandola, F.; Balzani, V. J. Chem. Educ. 1983, 60, 814.
Sun, L.; Hammarström, L.; Norrby, T.; Berglund, H.; Davydov, R.; Andersson, M.;
Börje, A.; Korall, P.; Philouze, C.; Almgren, M.; styring, S.; Åkermark, B. J. Chem.
Soc. Chem. Commun. 1997, 607.
Sun, L.; Berglund, H.; Davydov, R.; Norrby, T.; Hammarström, L.; Korall, P.; Börje,
A.; Philouze, C.; berg, K.; Tran, A.; Andersson, M.; Stenhagen, G.; Mårtensoon, J.;
Almgren, M.; Styring, S.; Åkermark, B. J. Am. Chem. Soc. 1997, 119, 6996.
48
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
Magnuson, A.; Frapart, Y.; Abrahamsson, M.; Horner, O.; Åkermark, B.; Sun, L.;
Girerd, J.-J.; Hammarström, L.; Styring, S. J. Am. Chem. Soc. 1999, 121, 89-96.
Sjödin, M.; Styring, S.; Sun, L.; Åkermark, B.; Hammarström, L.; J. Am. Chem. Soc.
2000, 122, 3932-3936.
Berg, K. E.; Tran, A.; Raymond, M. K.; Abrahamsson, M.; Volny, J.; Redon, S.;
Andersson, M.; Sun, L.; Styring, S.; Hammarström, L.; Toftlund, H.; Åkermark, B.
Eur. J. Inorg. Chem. Accepted 2000.
Sun, L.; Raymond, M. K.; Magnuson, A.; LeGourriérec, D.; Tamm, M.;
Abrahamsson, M.; Huang Kenéz, P.; Mårtensson J.; Stenhagen, G.; Hammarström, L.;
Styring, S.; Åkermark, B. J. Inorg. Biochem. 2000, 78, 15-22.
Demas, J. N.; Addington, J. W. J. Am. Chem. Soc. 1976, 98, 5800
Gaines, G. L.; O´Neil, M. P.; Svec, W. A.; Niemczyk, M. P.; Wasielewski, M. R. J.
Am. Chem. Soc. 1991, 113, 719.
Wasielewski, M. R.; Johnson, D. G.; Svec, W. A.; Kersey, K. M.; Minsek, D. W. J.
Am. Chem. Soc. 1988, 110, 7219.
Chen, P.; Meyer, T. J. Inorg. Chem. 1996, 35, 5520.
Bock, C. R.; Meyer, T. J.; Witten, D. G. J. Am. Chem. Soc. 1975, 97, 2909.
Serpone, N. In Photoinduced Electron Transfer; Fox, M. A., Chanon, M., Eds.;
Elsevier: Amsterdam, 1988; Vol. Part D, p 47.
Darwent, J. R.; Kalyanasundaram, K. J. Chem. Soc. Faraday Trans. 1981, 77, 373.
Land, E. J.; Prütz, W. A. Int. J. Radiat. Biol. 1979, 36, 75.
Babcock, G. T.; Sauer, K. Biochim. Biophys. Acta 1975, 376, 315.
Barry, B. A.; Babcock, G. T. Chem. Scr. 1988, 28 A, 117.
Burdinski, D.; Wieghardt, K.; Steenken, S.; J. Am. Chem. Soc. 1999, 121, 10781.
Burdinski, D.; Bothe, E.; Wieghardt, K. Inorg. Chem. 2000, 39 105.
Sun, L.; Hammarström, L.; Åkermark, B.; Styring, S. Chem. Soc. Rev. 2001, 30, 3649.
Yonemoto, E. H.; Riley, R. L.; Kim, Y.; Atherton, S. J.; Schmehl, R. H.; Mallouk, T.
E. J. Am. Chem. Soc. 1992, 114, 8081-8087.
Goulle, V.; Harriman, A.; Lehn, J.-M. J. Chem. Soc. Chem. Commun. 1993, 1034.
Berkoff, R.; Krist, K; Gafney, H. D. Inorg. Chem. 1980, 19, 1.
Opperman, K. A.; Mecklenburg, S. L.; Meyer, T. J. Inorg. Chem. 1994, 33, 52955301.
Treadway, J. A.; Chen, P.; Rutherford, T. J.; Keene, F. R.; Meyer, T. J. J. Phys. Chem.
1997, 101, 6824-6826.
Larson, S. L.; Elliot, C. M.; Kelley, D. F. J. Phys. Chem. 1995, 99, 6530.
Schlicke, B.; De Cola, L.; Belser, P; Balzani, V. Coord. Chem Rev. 2000, 208, 267.
Serroni, S.; Juris, A.; Venturi, M.; Campagna, S.; Resino, I.; Denti, D.; Credi, A.;
Balzani, V. J.Mater. Chem. 1997, 7, 1227-1236.
49
[72]
[73]
[74]
[75]
[76]
[77]
[78]
Denti, G.; Campagna, S.; Sabatino, L.; Serroni, S.; Ciano, M.; Balzani, V. Inorg.
Chem. 1990, 29, 4750-4758.
Denti, G.; Serroni, S.; Campagna, S.; Ricevuto, V.; Balzani, V. Inorg. Chim. Acta.
1991, 182, 127-129.
Campagna, S.; Denti, G.; Serroni, S.; Ciano, M.; Balzani, V. Inorg. Chem. 1991, 30,
3728-3732.
Juris, A.; Balzani, V.; Campagna, S.; Denti, G.; Serronni, S.; Frei, G.; Güdel, H.
Inorg. Chem. 1994, 33, 1491-1496.
Campagna, S.; Denti, G.; Sabatino, L.; Serroni, S.; Ciano, M.; Balzani, V. J.Chem.
Soc., Chem. Commun. 1989, 1500.
Balzani, V. unpublished results
Meyer, G. personal communication
50